Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 15-GEMMASCOPE-RES-16K
    6. 2114
    Prev
    Next
    INDEX
    Explanations

    instances of the word "For" and its variations, indicating a focus on phrases that introduce reasoning or information

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ConstraintMaker
    -0.97
     ་་
    -0.83
    IonicModule
    -0.81
    DotNetBar
    -0.81
    ImageContext
    -0.79
    ."],
    -0.77
    URLException
    -0.75
    NUMX
    -0.75
    '};
    -0.74
     esternos
    -0.73
    POSITIVE LOGITS
     For
    0.91
    For
    0.90
     FOR
    0.74
     purposes
    0.66
     example
    0.63
     for
    0.62
    FOR
    0.62
    for
    0.58
     instance
    0.58
     reasons
    0.58
    Activations Density 0.139%

    No Known Activations